- Email: [email protected]
Thermal comfort of patients: Objective and subjective measurements in patient rooms of a Belgian healthcare facility
Building and Environment 46 (2011) 1195e1204
Contents lists available at ScienceDirect
Building and Environment journal homepage: www.elsevier.com/locate/buildenv
Thermal comfort of patients: Objective and subjective measurements in patient rooms of a Belgian healthcare facility Jan Verheyen a, *, Nele Theys a, Luk Allonsius a, Filip Descamps b,1 a b
University College of Antwerp, Department of Applied Engineering & Technology, Construction Engineering, Paardenmarkt 92, 2000 Antwerpen, Belgium Vrije Universiteit Brussel, Department of Architectural Engineering, Pleinlaan 2, 1050 Brussel, Belgium
a r t i c l e i n f o
a b s t r a c t
Article history: Received 17 September 2010 Received in revised form 27 November 2010 Accepted 11 December 2010 Available online 21 December 2010
In healthcare facilities, the prediction of mean thermal comfort perception of patients and staff is necessary to formulate requirements for the architectural and building systems design and control, and for establishing guidelines for the use of clothing and bedding systems. In this study thermal comfort of patients is evaluated by comparing objective (environmental and personal) parameters and subjective measurements (questionnaires) of thermal comfort for different groups of patients, according to the ward they are occupied in. The study involved 99 patients of maternity, oncology, neurology, gastro-enterology, abdominal surgery and thoraco-vascular surgery wards. T tests reveal no significant difference between Predicted Mean Vote (PMV) obtained from objective measurements and Actual Mean Vote (AMV) for all the different wards except for neurology. Binomial tests show that the difference between the predicted percentage of dissatisfied (PPD) obtained by application of the PPD-formula in ISO 7730 as function of Actual Mean Vote and PPD obtained from personal acceptability votes is not significant on a 5% level for all wards. This means that PMV and PPD indices may be used to adequately predict mean thermal responses for these wards except for neurology. Ó 2010 Elsevier Ltd. All rights reserved.
Keywords: Thermal comfort Patients Hospital Healthcare facility
1. Introduction Design and operation of patient rooms should primarily aim at providing a healthy and healing environment for the patients recovering from surgery, injury or disease. There has been growing scientific evidence that the physical environment has an impact on health and well-being [1]. Every physiological strain applied to the patient will induce extra stress on top of stress related to the disease or injury of the patient which is undesired unless medical treatment requires so. The thermal environment can also be an important source of undesired physiological strain on the body. It is therefore necessary to know how patients perceive thermal comfort. Thermal comfort of patients can be different from that of the healthy population because the nature of the physical disability will affect thermophysiology, thermal sensation, metabolism, blood flow, regulatory response, such as vasomotor control of body skin temperature or the ability to sweat. Depending on the disability and health status of a patient, the adaptive opportunity may be
* Corresponding author. Tel.: þ32 (0) 494 9113 52 (Mobile); fax: þ32 (0) 3 23186 70. E-mail address: [email protected] (J. Verheyen). 1 Daidalos Peutz bouwfysisch ingenieursbureau. 0360-1323/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.buildenv.2010.12.014
restricted, limiting the range of thermal environmental conditions providing comfort. Also the use of technical aids (such as wheelchairs), medical treatment and the use of drugs can affect thermal sensation and comfort perception. The aged are more largely represented in hospitals in Belgium than they have part in the overall demographic composition. It is generally considered that the ability to thermoregulate decreases with age [2] and that older people prefer slightly higher temperatures [3] although the latter can be attributed to lower activity levels and clothing differences of the elderly compared to their younger counterpart.
1.1. Guidelines and recommendations for thermal environmental conditions of patient rooms Some recommendations and guidelines for design of patient rooms for healthcare inpatient nursing are summarized in Table 1. According to Ref. [5] each room should have individual temperature control. Lower or higher temperatures than the ones prescribed in the standard shall be permitted when patients’ comfort and (or) medical treatment require those conditions [4]. Thermal comfort requirements should be considered on an individual basis for people with physical disabilities [3]. This is confirmed in the international standard for thermal comfort for
1196
J. Verheyen et al. / Building and Environment 46 (2011) 1195e1204
Table 1 Thermal comfort ranges in some standards and guidelines. RH [%]
Design temperature [ C]
Maximum air velocity [m/s]
ASHRAE; 2008 [4] ASHRAE; 2007 [5]
Max. 60 30 (W)e50 (S)
21e24 21e24
e e
ISIAQ; 2003 [6] College Bouw ziekenhuis-voorzieningen; 2002 [7] VIPA; 2002 [8]
<60%, preferable <50% e
e 22.5e25.5
e e
40e65 30e70 30e70
23.5e25.5/21e23 23e26/20e24 22e27/19e25
0.18/0.15 0.22/0.18 0.25/0.21
Inpatient nursing: patient room Hospital and outpatient facilities: nursing: patient room Healthcare: patient rooms Uses 0.5 < PMV < 0.5 for acceptable comfort range Optimal (summer/winter) Normal (summer/winter) Minimal (summer/winter)
(W, S) ¼ Winter, Summer.
people with special requirements [9], stating that consideration of individual requirements may be even more necessary for the disabled than for standard persons. Furthermore, the inclusion of personal control possibilities is recommended [10].
1.2. Previous research For evaluation of thermal comfort and prescriptions of design requirements, models based on Fanger [11] are worldwide most frequently used in current standards to assess thermal comfort, in some cases improved by an adaptive thermal comfort model for naturally ventilated buildings [12]. These models, however, are only applicable to healthy adults in moderate thermal environments, because they were based on laboratory and field studies involving mainly college-aged students in moderate environments. ISO 14415 [9] is meant for application in addition to ISO 7730 [13] for determination and interpretation of thermal comfort in the case of people with special requirements, such as people with physical disabilities, pregnancy, sickness or normal aging. This standard contains general considerations concerning thermal comfort perception as well as specification of conditions for thermal comfort. Also changes in thermophysiology and thermosensation, thermoregulatory responses or thermal comfort perception are discussed and summarized for some specific disabilities in relation to thermal environments of concern. 1.2.1. Adaptive thermal comfort and patients An adaptive thermal comfort theory was developed as a result of extensive field studies that revealed non-universality of the applicability of traditional thermal comfort theory based on laboratory experiments. Adaptation can be interpreted broadly as the gradual diminution of the organism’s response to repeated environmental stimulation. More specific, adaptation can be categorized in three main groups [12]: behavioral adjustment (personal, technological and cultural responses that change the heat and mass flows reigning the thermal balance of the human body), physiological
Table 2 Total clothing insulation IT [Clo] taken from Ref. [30]. Ac (%) Percentage of body surface area coverage: 23.3 48.0 59.1 67.0 94.1 M1 M1 M1 M1 M1 M1 M1 M1 M1
þ þ þ þ þ þ þ þ þ
Q1 þ S1 Q2 þ S1 B þ S1 Q1 þ S2 Q2 þ S2 B þ S2 Q1 Q2 B
1.57 1.57 1.57 1.38 1.38 1.38 0.98 0.98 0.98
2.15 1.84 1.82 1.65 1.53 1.43 1.16 1.14 1.07
2.72 2.24 2.08 2.15 1.93 1.76 1.43 1.42 1.24
2.88 2.41 2.18 2.62 2.20 1.80 1.90 1.69 1.45
4.56 3.73 2.56 4.34 3.55 2.40 4.03 3.03 2.11
adaptation, resulting from exposure to thermal environments in diminution of the strain induced by this exposure and psychological adaptation, including the altering of perception of sensor signals due to past experiences and future expectations that in turn alters reaction of the human body in order to maintain thermal equilibrium. The PMV/PPD is a static heat balance model that can adequately predict mean thermal sensation votes for indoor environments that occupants have no or limited control over, such as centrally controlled HVAC buildings. It can be seen as a partially adaptive thermal comfort model since it incorporates the effects of some behavioral adaptations to thermal environment like clothing and air velocity adjustments. Adaptive thermal comfort theory incorporates all effects of the opportunity to adapt. It allows a broader range for indoor acceptable environments as function of outdoor thermal environmental characteristics and is optional in ASHRAE STD 55 [14], ISO 7730 [13] and EN 15251 [15] for occupant controlled, naturally conditioned buildings with operable windows in warm periods or warm climatic zones. People with physical disabilities have restricted adaptive opportunity and special attention should be paid to this user group especially in conditions away from thermal neutrality as uncomfortable conditions affect patients both physically and mentally [16]. Patients may have restricted mobility and the ability to thermoregulate by behaving appropriately (e.g. move out of non comfortable environment or adjust clothing) may be severely restricted. Therefore the adaptive capability should be assessed when designing for comfortable indoor environment [17]. 1.2.2. Laboratory studies A reliable method to obtain ranges of acceptable thermal comfort conditions, is to conduct thermal comfort enquiries in combination with objective measurements of physical environmental and personal parameters of a group of people representative for total population of interest in a controlled climatic chamber. For patients with particular disabilities, these tests were conducted by Parsons and Webb [3,17,18] and by Yoshida et al. [19]. Of course, for reasons of medical treatment, these experiments cannot be conducted for (a representative group of) all patients. In an exploratory laboratory experiment, Webb and Parsons compared objective and subjective measurements of 16 persons with and without physical disabilities [20]. The disabilities concerned included cerebral palsy, spina bifida, stroke, Friedrich’s ataxia, blindness, paralysis, heart condition, encephalitis, Guillain-Barré syndrome, missing limbs and metal in legs. The results of laboratory experiments show no significant differences (p < 0.05) between mean responses of subjects with and without physical disabilities. Variation of responses was greater in cool to neutral conditions and smaller for slightly warm to warm conditions. The authors concluded
J. Verheyen et al. / Building and Environment 46 (2011) 1195e1204
that thermal comfort requirements for people with physical disabilities should be considered on an individual basis. Results of the main laboratory studies were integrated into a software tool that allows to determine the thermal comfort requirements of people with physical disabilities [18]. General guidance is provided in terms of a summary of findings as well as mean and variation in responses and the possibility to interrogate individual responses. The software incorporates results from people with cerebral palsy, spinal injury, accident including tetra- and quadriplegia, spinal degenerative conditions, spina bifida, hemiplegia including stroke, polio, osteo-arthritis, rheumatoid arthritis, head injury and multiple sclerosis. The laboratory results show that ISO 7730 [13] can be used without modification as a first approximation to the average responses of people with and without physical disabilities. The variation in responses is different from that of the healthy population and individual characteristics should be considered when assessing thermal environment or determining requirements for people with physical disabilities. 1.2.3. Field studies Apart from these laboratory studies, several field studies were conducted concerning thermal comfort of people with different physical disabilities, using different methods. Hill et al. [3,16] conducted field studies of people with physical disabilities and performed enquiries of carers and professionals showing that functional abilities diminish for people with physical disabilities when thermal environments are not correctly targeted at their needs, which may lead to discomfort, increased pain and poorer quality of life. In 4 Iranian hospitals, objective measurements of thermal comfort perception of patients in 14 hospital rooms are compared with requirements of thermal comfort in standards [21]. In another study, the possibilities to reconcile the different expectations towards thermal comfort of patients and staff in the same room are compared [22]. The authors suggest different zones for patients and for staff because of different thermal comfort requirements [22]. This can be achieved by differentiating radiant temperature by application of directional radiant heating via warm surfaces or ceiling for patients and cooling via the floor for staff. Attention on ventilation is needed to provide low absolute humidity levels to avoid condensation on the cooling floor. In an orthopaedic ward in a Swedish hospital the thermal environment was analyzed using objective measurements in combination with subjective measurements. Forty non-bedridden patients and 35 staff members were questioned in winter and summer season [23]. The difference between patient and staff thermal comfort perception is greater in winter than it is in summer despite similar temperatures in both seasons. For both seasons considered separately, calculated values of temperature for acceptable thermal comfort differ for patients and staff, in spite of both groups indicating to accept the thermal environment in the questionnaires. It was concluded that staff and patients cannot be treated as one coherent group of users with the same needs and preferences concerning indoor environment. The same data, but pooled for patient and staff population, were used to investigate the usability of objective and subjective indicators of indoor overall comfort by multiple regression analysis with physical and rated indicators apart and together [24]. Subjective sensory ratings were significantly better than objective indicators at predicting overall rated indoor comfort. The best prediction result of overall rated indoor comfort was obtained using a combination of subjective measures and physical variables. In Taiwan, one of the most extensive field studies of thermal comfort conducted in a hospital environment concerning patient thermal comfort perception is conducted [25]. A comparison of
1197
thermal environment of the hospital was made with ASHRAE standard 55 recommendations on one hand and a comparison with subjective measurements on the other hand. Probit regression analysis resulted in preferred effective temperature (ET*) limits for the patient population. Based on chi-square tests, the authors concluded that physical strength has a highly significant effect on thermal sensation, especially in winter, but gender, age and acclimatization have not. Patients expected a warmer indoor environment than healthy people did. Preferred and neutral temperature for patients were 23.9 and 22.9 C ET* in winter and 24.6 and 24.0 C ET* in summer. The frail wanted a neutral temperature of 0.3 C ET* higher than that of the vigorous population and there was a 1.5 and 0.8 C difference in preferred ET* for winter and summer respectively. Regression model of AMV of patients on ET* indicated that patients thermal sensitivity is blunt. A study of thermal comfort was conducted in four hospitals in tropical Malaysia involving a total of 114 occupants [26]. Subjective and objective measurements were conducted simultaneously. However, only staff was involved in answering to the questionnaires. PMV calculated from objective measurements was compared to AMV of staff. The authors concluded that the significant deviation between AMV and PMV strongly implied that PMV method might not be suitable for application in hospitals in the tropics. 1.2.4. Different hospital users Studies have been done on the thermal comfort requirements for different hospital users [7,10,22]. This is done by considering the differences in thermal insulation of clothing and activity levels of different user groups. The influence of disease, treatment or other influence specific to the hospital setting is neglected. Results show the difficulty to reconcile the needs concerning the thermal environment of the different hospital users. Most field studies do not differentiate between patients as function of the type of disability or treatment taken, but consider patients and staff together [24], patients and staff apart [23], staff only [26] or patients only [25]. Other studies focus on the thermal comfort requirements for people with specific disabilities. For a review of the latter, the reader is referred to Parson’s book on human thermal environments [3], in which a complete chapter is dedicated to thermal comfort for special populations including people with physical disabilities. However no study is known by the authors that focuses on thermal comfort of different groups of hospital patient population according to the ward or department this population is occupied in.
1.3. Objective This study concentrates on different groups of patients, sorted per department or ward they are occupied in. Reason for this is of a practical nature; designing for comfortable indoor environment could be specifically diversified per department or ward. The nature of the disability will determine requirements, as the physical disability may influence thermal sensation and perception. To address the specific thermal comfort needs of patients, the design of the bed, the room and the control measures for the thermal environment per department or ward can be diversified. This approach is considered useful because patients are grouped in specific wards depending on the disease or disability they are receiving treatment for. It is intended to provide a first indication of the usability of the Fanger PMV (Predicted Mean vote) and PPD (Predicted Percentage of Dissatisfied) indices for groups of patient population per ward (according to the surgical recovery or medical treatment they are obtaining).
1198
J. Verheyen et al. / Building and Environment 46 (2011) 1195e1204
Fig. 1. Ac: percentage of body surface area coverage taken from [30].
2. Research method 2.1. Thermal comfort and bedding For the determination of the clothing insulation of patients, only few studies take bedding insulation into account [7,10,22]. In the former study [7] it is not indicated how total clothing insulation values for bedding systems (taken from measurements performed by McCullough et al. [27]) were transformed for input as clothing insulation values in PMV calculation according to ISO 7730 [13]. Estimated clothing insulation values have large influence on the results of the calculation of acceptable thermal comfort [10]. A comprehensive description of a model to account for heat and moisture transfer through clothing is described by Voelker et al. [28]. Global thermal comfort requirements for different hospital users in different situations (defined by clothing insulation ranges for summer and winter conditions and activity level ranges) including patients in bed are described in earlier research [10]. A method to calculate PMV and PPD for sleeping environments [29] is based on the thermal comfort index of Fanger where PMV is related to the imbalance between the actual heat flow from the human body and the heat flow required to obtain thermal comfort in the same environmental conditions. This imbalance is referred to as the thermal load and results in the PMV value by multiplying it with a sensitivity coefficient. This sensitivity is assumed to be valid also for sleeping conditions. The relation between PMV and PPD was also assumed by Lin and Deng to remain valid for sleeping conditions. The application of the method described in ISO 7730 is recommended to be limited to metabolic rate levels between 46 W/m2 and 232 W/m2 (0.8 and 4.0 Met). Application by Lin and Deng assumes that the same model for thermal comfort prediction can be used for activity levels of 40 W/m2 (0.7 Met), resembling metabolic heat production during sleeping activity. Lin and Deng formulate the heat balance equation at the skin surface in terms of operative temperature instead of mean radiant temperature and air temperature separately. The energy balance equation of Fanger’s model is adapted for heat and moisture transfer from the skin to the environment by taking the total clothing insulation resistance into account. The total insulation resistance of a bedding ensemble is measured for different bedding systems in combination with sheets and clothing worn [30] and includes the resistance of the air layer at the surface of the bedding
system and the human body at locations not covered. The air layer resistance is not included in the clothing insulation value used in the ISO 7730 method. The total insulation resistance represents a resistance to heat (or moisture) transfer uniformly spread over the entire body surface resulting in the same total heat (or moisture) transfer as the non-uniformly spread ensemble of bedding system and clothing worn. For moisture transfer, one value for the total vapor permeation efficiency (im ¼ 0.38) was considered to be reasonably accurate, also for bedding systems. Together with the Lewis ratio (LR, being he/hc; division between the evaporative (he) and convective (hc) surface transfer coefficients) and the total insulation resistance to heat transfer (Rt), this provides an approximation of the total resistance to moisture transfer (Re,t) through the following equation: im $ LR ¼ Rt/Re,t [31]. Skin wettedness w (being the evaporative heat loss from the skin (Esk) divided by the maximum possible evaporative heat loss (Emax) [31]), was chosen to be 0.06, which corresponds to no regulatory sweating for normal conditions. Evaporative heat transfer occurs only due to moisture diffusion through the skin, which holds true for M W (metabolic heat production work delivered) smaller or equal to 1 Met or 58.2 W/m2. Comfort equation proposed by Lin and Deng for sleeping environments [29]:
tsk;req t0 0:06im LR psk;s pa MW ¼ þ þ 0:014Mð34 ta Þ Rt Rt þ 0:0173Mð5:87 pa Þ
where M ¼ metabolism [W/m2] W ¼ work delivered [W/m2] tsk,req ¼ mean skin temperature, required for thermal comfort [ C] ¼ 35.7e0.028(M W) to ¼ operative temperature [ C] Rt ¼ total resistance to heat transfer of a bedding system including air layer around (covered) body im ¼ total vapor permeation efficiency [ND] ¼ 0.38 LR ¼ Lewis ratio [K/kPa] ¼ 16.5 K/kPa psk,s ¼ saturated water vapor pressure at tsk [kPa] pa ¼ ambient water vapor pressure [kPa] ta ¼ ambient air temperature [ C] PMV can be derived from the comfort equation using the following relationship [29]:
J. Verheyen et al. / Building and Environment 46 (2011) 1195e1204
1199
Table 3 Bedding coverage in current study, related to Table 2. Bedding garment [current study]
Sheet
Comforter single
Sheet þ blanket/Melton
Comforter double
Related to Table 2
B
Q1
Q2
Q3
PMW ¼ ½0:303 expð0:036MÞ þ 0:028L ¼ aL where L ¼ thermal load on the body ¼ difference between left hand side and right hand side of comfort equation a ¼ sensitivity coefficient Laboratory measurements [27,30] provide values for the total insulation resistance of bedding systems including the air layer around a covered bed for different bedding systems occupied by a thermal manikin in supine position on a single bed. The bedding systems measured in the studies consist of different combinations of quilt or blanket, bed, mattress and sleepwear, in different configurations of body surface area coverage. Measurements by Lin and Deng were conducted maintaining a fixed manikin surface temperature in an environmental chamber with air temperature and mean radiant temperature of 22 C, relative humidity of 50% and air velocity 0.15 m/s. Measurements by McCullough were conducted maintaining fixed manikin surface temperatures and constant heating power of the manikin (by altering environmental chamber temperature.) Air velocity was kept below 0.1 m/s. The result of the nude manikin placed horizontally in the chamber gave the same result for both studies; 0.73 Clo (¼0.113 m2 K/W) (Fig. 1). In the current study total insulation values of bedding systems were estimated by visual inspection relating to the detailed descriptions and photos of mattress (only one type was included and is referred to as M1), sheets and blankets as well as sleepwear and body surface area coverage in the original publication of Lin and Deng [30]. In Table 4, S1 refers to full-slip sleepwear and is described as long-sleeved, 100% cotton with thickness 0.46 mm, whereas S2 refers to half-slip sleepwear specified as short-sleeved, 100% cotton with thickness 0.83 mm in the original work by Lin and Deng [30]. 2.2. Data collection Objective (physical) and subjective measurements of thermal comfort were gathered for different groups of hospital patient population in a clinical healthcare facility in Belgium (climate: temperate; mild winters, cool summers; rainy, humid, cloudy) in March 2009, in-between winter and springtime. Only patients in single and double rooms were involved. The study includes 99 patients of different wards (see Table 5) for which local ethical committee approval was obtained.
Table 4 Clothing insulation characteristics of patients [30,31]. Description of clothing used for visual observation
[Clo]
Related to Table 2 in case of patient in bed
Sleeveless short gown (thin) Sleeveless long gown (thin) Short-sleeved hospital gown Short-sleeved short robe (thin) Short-sleeved pajamas (thin) Long-sleeved long gown (thick) Long-sleeved, short wrap robe (thick) Long-sleeved pajamas (thick) Long-sleeved, long wrap robe (thick)
0.15 0.20 0.31 0.34 0.42 0.46 0.48 0.57 0.69
S2 S2 S2 S2 S2 S1 S1 S1 S1
Patients were all individually interviewed in their room based on a questionnaire according to ISO 10551 [32] extended with other questions, considered relevant for the research project. Supplementary to the basic questions suggested in ISO 10551, questions about the satisfaction concerning indoor air quality, acoustic quality, light and visual comfort, the stability of the thermal environment, the possibility to adapt and health status were added. Starting prior to the interview, air temperature, operative temperature, relative humidity and omnidirectional air velocity were measured at a height of 1.10 m in a zone of 1 m around the patients bed with a Brüel & Kjær indoor climate analyzer model number 1213. Personal factors; activity level and clothing insulation level were estimated based on answers to the questionnaire and on observation. Clothing and bedding thermal insulation values were derived from ASHRAE standard 55 [14] and measurements by Lin and Deng [30]; see Tables 2e4. PMV and PPD were calculated according to ISO 7730 [13] and to the method developed by Lin and Deng [29] (the latter in the case of patient lying in a bed). These measurements are referred to as objective measurements. The results of the objective measurements are then used as a reference indicating the predicted mean thermal sensation responses of healthy adults. 2.3. Data analysis Two-sample t tests were conducted for thermal sensation to determine whether the difference between the mean of PMV obtained from objective and from subjective measurements is significant on a 5% level. Binomial tests were used to reveal whether the difference of the means of predicted percentage of dissatisfied obtained from objective and subjective measurements was significant on a 5% level. The objective PPD is obtained from the mean of subjective thermal sensation votes by calculation of PPD according to ISO 7730 [13]. The subjective PPD is calculated from personal acceptability votes. These data analysis methods are in agreement with the methods described in ISO 10551 [32]. 3. Results and discussion 3.1. Personal parameters Observations of total thermal insulation IT and clothing thermal insulation Icl and activity level of patients are represented by boxplots in Fig. 2. Almost half of all patients (46/99) had been either reclining or sleeping in bed during the hour prior to enquiry. Almost one-third of all patients (31/99) was just sitting or reading or writing in Table 5 Departments, wards and number of patients involved. Department Maternity Medicine
Surgery Total
Ward
Oncology Neurology Gastro-enterology Abdominal surgery Thoraco-vascular surgery
Number of patients involved 17 17 17 14 17 17 99
1200
J. Verheyen et al. / Building and Environment 46 (2011) 1195e1204
Fig. 2. Activity level and bedding and/or clothing insulation of patients.
100% 90%
maternity (mean 0.06)
80%
medicine oncology (mean -0.24) medicine neurology (mean -0.24)
70%
medicine gastro-enterology (mean -0.14)
60%
abdominal surgery (mean -0.65)
number 50% of votes
thoracovascular surgery (mean -0.35) total (mean -0.26)
40% 30% 20% 10% 0% very weak
weak
neutral
fit
very fit
Self-described health status Fig. 3. Self-described health status.
a chair, the rest was walking about and one patient had been mostly standing, relaxed during the last hour. This distribution in activity level was similar for all wards, but slightly different for oncology and abdominal surgery wards; only one of all oncology patients had been walking about and only 5 out of 17 abdominal surgery patients had been either reclining or sleeping in bed . The insulation provided by clothing and bedding system is represented by the boxplot in the middle of Fig. 2. The insulation provided by clothing of the patients that are standing or sitting in a chair is represented by the right hand side part of Fig. 2. The difference between parameters total insulation IT and clothing insulation Icl is explained in Section 2.1 of this document. In case of a patient sitting in a chair, the insulation characteristics of the chair were neglected. A sitting posture results in a decreased insulation of clothing due to compression of air layers in the clothing [14]. This
decrease may be offset by the insulation provided by the chair. For many chairs the net effect of sitting is a minimal change in clothing insulation. For this reason, ASHRAE standard 55 recommends that no adjustment to clothing insulation is to be made if there is uncertainty as to the type of chair [14]. Since both PMV/PPD methods are valid only for steady state conditions, a supplementary question was added in the questionnaire, enabling exclusion from the results when substantial altering of thermal conditions of the environment in the previous hour was indicated by the patients’ answer. No altering of thermal environment was reported in the questionnaires. Physical strength (answers to questions: do you feel very, moderate, slightly frail or vigorous) has a highly significant effect on a 5% level (p ¼ 0.000) on thermal sensation [25]. Therefore, the authors found it useful to conclude a question about the health
Table 6 Results and environmental parameters for patients grouped per category of self-described health status. Self-described health status
Number of observations
Mean PMVobj
Mean PMVsubj
STD PMVobj
STD PMVsubj
Mean operative temperature [ C]
STD operative temperature [ C]
Mean Relative Humidity [%]
STD Relative Humidity [%]
Very weak (2) Weak (1) Neutral (0) Fit (þ1) Very fit (þ2)
7 33 40 17 2
0.32 0.76 0.40 0.58 0.49
0.57 0.00 0.15 0.53 2.00
1.27 0.99 1.13 0.95 0.68
1.72 1.22 0.83 1.07 0.00
24.0 23.4 23.7 23.3 24.1
0.9 0.9 1.0 0.5 2.3
37.0 36.4 35.7 36.1 33.5
3.8 4.4 4.5 3.1 10.6
STD ¼ standard deviation.
J. Verheyen et al. / Building and Environment 46 (2011) 1195e1204
1201
100% maternity (mean 1.12)
90%
medicine oncology (mean 0.71)
80%
medicine neurology (mean 0.94)
70%
medicine gastro-enterology (mean 0.71) 60% abdominal surgery (mean 0.29) number 50% of votes 40%
thoracovascular surgery (mean 0.88) total (mean 0.78)
30% 20% 10% 0% very unsatisfied unsatisfied
neutral
satisfied
very satisfied
possibilities to adapt Fig. 4. Satisfaction with possibilities to adapt to thermal environment.
status of the patients in the enquiry. Results of self-described health status are represented in Fig. 3. Fig. 3, depicting percentages of answers in the categories of selfdescribed health status shows a clear deviation to the negative side (weak and very weak), which is to be expected in a healthcare facility. Table 6 contains mean and standard deviation of objective PMV and subjective PMV together with environmental data of operative temperature and relative humidity of the rooms of all the patients grouped per self-described health status category. Results of linear
regression of subjective PMV on self-described health status show the correlation coefficient squared to be 0.066, indicating that there is indeed an effect of influence that cannot be ignored. However, Table 6 shows that patients of one group voting in the same category of health status did not experience the same physical conditions as another group. Mean objective PMV for different groups is between 0.32 and 0.76. Nevertheless, if one would wish to develop a method for determining the PMV for overall patient population and is looking for parameters of influence to predict the mean thermal sensation vote, one could consider incorporating the
Fig. 5. Operative temperature (To), relative humidity (RH) and air velocity in hospital patient rooms.
Fig. 6. Acceptable thermal environment for all measurements.
1202
J. Verheyen et al. / Building and Environment 46 (2011) 1195e1204
Table 7 Results of objective and subjective measurements, two-sample t tests and binomial tests. 5% level of significance
Number of Mean Mean STD STD Difference between PMVobj Mean Mean Difference between observations PMVobj PMVsubj PMVobj PMVsubj and PMVsubj PPDobj PPDsubj PPDobj and PPDsubj* objective or (binomial test) subjective
Maternity Maternity Medicine Oncology Neurology Gastro-enterology Surgery Abdominal surgery Thorovascular surgery
17 17 17 14 17 17
0.30 0.58 0.87 0.58 0.45 0.52
0.29 0.06 0.06 0.21 0.12 0.41
0.93 1.00 1.09 1.01 0.94 1.35
0.99 1.20 1.09 1.63 0.93 1.00
Not significant (p ¼ 0.082) Not significant (p ¼ 0.101) Significant (p ¼ 0.038) Not significant (p ¼ 0.135) Not significant (p ¼ 0.085) Not significant (p ¼ 0.787)
6.8 5.1 5.1 6.0 5.3 8.5
5.9 12.0 0.0 7.1 0.0 5.9
Not Not Not Not Not Not
significant significant significant significant significant significant
(p (p (p (p (p (p
¼ ¼ ¼ ¼ ¼ ¼
0.680) 0.212) 0.413) 0.576) 0.396) 0.568)
*PPDsubj calculated from personal acceptability votes. STD ¼ standard deviation.
effect of health status in a model for PMV. If very weak, weak, neutral, fit and very fit would respectively correspond to 2, 1, 0, 1 and 2, the mean self-described health status has a value of 0.26; in between weak and neutral. There is however no indication that the categories used are psychologically interpreted to have the same distance from each other. There is one population group that reported a positive mean self-described health status; namely maternity. Abdominal surgery shows most negative of mean values. The mean for each group is indicated in the legend of Fig. 3. Fig. 4, shows the satisfaction (on a five-point scale) of the patients concerning possibilities to adapt to the thermal environment when they were having too cold or too hot. As examples, opening windows, changing temperature, altering bedding or clothing, sunblinds, ventilation,., were included in the questionnaire. Overall patients were satisfied with their possibilities to adapt to the thermal environment. If 2 is very unsatisfied, 1 is unsatisfied, 0 is neutral, þ1 is satisfied and þ2 is very satisfied, overall patients mean is 0.78; in between neutral and satisfied. The mean for each group is indicated in the legend of Fig. 4. There is however no indication that the categories used are psychologically interpreted to have the same distance from each other. Apart from their satisfaction, 68% of patients indicated they were able to adjust to the thermal environment by altering clothing or bedding by themselves and needed no help from nursing staff. This means 32% needed help to do so. 54% indicated being able to adjust to thermal environment by other measures (like opening windows, heating, cooling,.) without help from nursing staff. Most self-control in possibility to adapt by altering clothing or bedding as well as by other measures was found in maternity patients and least in abdominal surgery patients. Not a single patient indicated not being able to adjust. However, the need for help means an extra obstacle in achieving the adjustment. The patient has to rely upon judgement and assistance of a third person. First of all the posing of
the question implies admitting insufficient self-control and independence. Secondly, the nursing staff at service should be convinced of the validity of the complaint before taking action to adjust. Very likely the perception of thermal comfort of staff is different from that of the patient [10]. In the case of thermal discomfort these elements can lead to a delay or even cancellation of the adjustment and risks the patient of drifting further away from comfortable conditions. Nevertheless, patients are rather satisfied with their possibilities to adapt, with a deviation towards neutral for abdominal surgery patients. Maternity patients show to be most satisfied with their possibilities to adapt (Fig. 5). 3.2. Indoor environmental parameters The measurements of indoor parameters operative temperature and relative humidity in patient rooms are compared with design parameters recommended by ASHRAE and VIPA (Table 1). Temperature in the rooms is high, certainly for winter period, and some rooms fail to meet the lower limit for relative humidity level compared to VIPA and ASHRAE recommendations (Table 1). The rooms with lowest relative humidity and highest operative temperature are mainly oncology rooms. Lowest relative humidity levels indoors occur in wintertime when no humidification is provided. Air velocity for optimal design level is limited to below 0.15 m/s in wintertime according to VIPA [8]. Only two measurements (of a total of 99) trespass this limit and only one measurement (0.5 m/s) is higher than the maximum value for minimal comfort level of 0.21 m/ s set by [8]. Augmenting air velocity from 0.1 m/s to 0.2 m/s approximately corresponds to a decrease in operative temperature of 0.9 C or a decrease in PMV value of 0.33 (for sedentary activity (1 Met) in summer clothing (0.5 Clo) at 50% RH). The high air velocities measured do not occur in rooms with high room temperature.
Fig. 7. Boxplots of PMV obtained from objective and subjective measurements of different wards’ population.
J. Verheyen et al. / Building and Environment 46 (2011) 1195e1204
Fig. 6 shows the percentage of total measurements of thermal environments that are acceptable. The left hand side of Fig. 6 shows percentage of acceptable thermal environments based on 3 different determination methods; percentage of votes in central three categories of thermal sensation, percentage of votes of personal acceptability in category ‘acceptable’ and percentage of votes of thermal preference in category ‘want no change’. On the right hand side of Fig. 6 the percentage of environments that meet the criteria of acceptable thermal comfort as indicated in Table 1 is represented. In all cases operative temperature, relative humidity and air velocity measurement results (the latter only for VIPA recommendations) are checked with the ranges for acceptable thermal comfort. It can be seen that although 95% accepted the thermal environment they were in on a direct assessment, only 74% wanted no change. Overall, patients in this healthcare facility found the environment acceptable rather than unacceptable. The unacceptability index (percentage of votes in category ‘unacceptable’) was similar for all wards with an overall value of 5%. Only for oncology patients it was higher; 12% and for neurology and abdominal surgery wards there were no patients that judged the environment to be unacceptable. 78% voted in central three categories, which is almost the same as the number wanting no change. From objective measurements of operative temperature, relative humidity and air velocity, it can be concluded that 71% of environments comply with ASHRAE and 70% with VIPA for normal quality thermal environment in winter. Furthermore 9% of environments did not comply with VIPA minimal level of thermal environment. However, of all patients included in this study that were occupying these environments that do not meet the VIPA minimum criteria, only 22% voted outside the central 3 categories of thermal preference, only 33% indicated to want change and they all found it to be perfectly bearable. The central tendency (median) for answers to the tolerance assessment question (“Is it perfectly bearable/./unbearable?”) is for overall patient population and for all wards separately the category ‘perfectly bearable’. 11% of overall patient population judged it to be more or less difficult to tolerate with highest painfulness indices (percentage of judgements expressing ‘difficulty to bear’) in wards oncology (18%) and maternity (18%), gastro-enterology (14%) and lowest in neurology (6%), abdominal surgery (6%) and thoraco-vascular surgery (6%). Only one patient judged it to be unbearable. This patient was occupied in the maternity ward. 3.3. PMV/PPD indices Values in Table 7 for mean PMVobj, STD PMVobj and PPDobj and boxplots of all objective measurements in Fig. 7 are based on data of individual PMV value calculation by the method according to ISO 7730 [13] in case of patient not in bed and by the method developed by Lin and Deng [29] in case of patient in bed. Fig. 7 and Table 7 show that for all groups patients’ rated thermal sensation from subjective measurement is higher than the PMV obtained from objective measurements. This indicates that patients expect a lower thermo-neutral temperature in comparison to healthy adults. This is in contradiction with results obtained in a previous field study [25] conducted in a hospital be it in a very different climatic zone. Apart from outdoor climate, another reason for this difference could be found in the fact that patients voted very often in the neutral category and showed a high acceptability rate, indicating a bluntness to thermal sensation in comparison to healthy population. This was also observed and reported in previous research [25]. Another reason for the difference between the objective and subjective PMV, can be a systematic underestimation of thermal insulation of clothing and bedding system and/or of metabolic heat production, resulting in a lower objective PMV. Variance in PMV
1203
obtained from subjective measurements is slightly bigger than variance in PMV from objective measurements for all wards except for thoraco-vascular surgery. The largest difference is found for the gastro-enterology ward. Both laboratory and field studies show a wide variation in thermal responses of the disabled [9]. A bigger variance in PMV obtained from subjective measurements in comparison to those obtained from objective measurements is therefore not surprising. Two-sample t tests show that the difference between PMV from subjective measurements (AMV) and PMV from objective measurements is not significant on a 5% level for all individual ward population groups except for neurology. A second analysis was performed using the same techniques but after rejection of outliers determined with Grubbs’ test with a significance level of 5%. The conclusions remain the same. Finally, a third analysis is conducted using Welch’s t-test, assuming unequal variances for results of objective and subjective measurements. This leads to smaller number of degrees of freedom and slightly smaller t values. The results and conclusions remain the same, only with slightly different p-values. The p-values given in Table 7 correspond to Welch’s t-test results. Binomial tests reveal no significant difference between PPD obtained from subjective measurements and PPD calculated from subjective thermal sensation votes for all wards individually. This means that for wards’ populations included in this study, calculation method for PPD according to ISO 7730 [13] as function of PMV can be used to calculate the predicted percentage of dissatisfied for patient population of all wards concluded in this study. 4. Conclusions Thermal environment in this Belgian healthcare facility is acceptable for 95% of the patients (direct assessment), despite of 29% of thermal environments not complying with ASHRAE design ranges of temperature and relative humidity. This could indicate that environmental conditions for hospital patient rooms recommended by ASHRAE are too stringent for the patients included in this study or that adaptation possibly allows broader regions of recommended environmental parameters. PMV of patients can be calculated using ISO 7730 [13] and using a calculation method described by Lin and Deng [29] (the latter in case of the patient lying in bed) for all wards included in this study except for neurology ward. PMV may adequately predict mean thermal sensation for the majority of patient population for the wards considered except for neurology ward. The mean PMV of patients occupied in neurology ward is significantly different from the mean PMV of healthy population on a 5% level. Neurology ward occupies patients that need medical treatment for, or that are recovering from injury or surgery of the brain, bone marrow or nervous system. These parts are of vital importance to human thermophysiology and eregulation and it is therefore not surprising that thermal comfort perception of these patients differs significantly from that of the healthy population. Patients rated their thermal sensation systematically higher than predicted by objective measurements, indicating that thermoneutral temperature for patients is lower than for healthy, young population. This is not in agreement with results of a previous field study [25]. However only 5% of patients judged the thermal ambiance as unacceptable on a personal level. The higher mean of sensation votes could be the result of thermal sensation bluntness of patient population or of systematically underestimated clothing insulation and/or metabolic heat production resulting in an underestimated objective PMV. The difference between the PPD calculated by use of ISO 7730 [13] from PMV obtained from the questionnaires and PPD directly obtained from personal acceptability answers in the questionnaires
1204
J. Verheyen et al. / Building and Environment 46 (2011) 1195e1204
is not significant on a 5% level for all wards’ populations. PPDformula taken from ISO 7730 [13] as function of PMV can be used for patient population considered in this study. Acknowledgements This research project was conducted in collaboration with Ziekenhuis Oost-Limburg (ZOL). The authors wish to thank ZOL (especially Tilly Postelmans and employees of the quality control division and Philip Verheye of facility management division) for enabling the execution of the project and Jacques Geers of University College of Antwerp for assistance in the statistical analysis. References [1] Van den Berg AE. Health impacts of healing environments e A review of evidence for benefits of nature, daylight, fresh air, and quiet in healthcare settings. 2005. ISBN: 90 327 03447. [2] Havenith G. Temperature regulation and technology. Gerontechnology 2001;1 (1):41e9. [3] Parsons KC. Human thermal environments: the effects of hot, moderate and cold environments on human health, comfort and performance. 2nd ed. London: Taylor and Francis; 2003. [4] ASHRAE/ASHE STD 170-2008. Ventilation of health care facilities. Atlanta, GA: American Society of Heating, Refrigerating, and Air-conditioning Engineers; 2008. [5] ASHRAE. Handbook HVAC applications: chapter 7; healthcare facilities. Atlanta, GA: American Society of Heating, Refrigerating, and Air-conditioning Engineers; 2007. [6] ISIAQ. Indoor air quality for healthcare facilities. International Society of Indoor Air Quality and Climate; 2003. [7] College Bouw Ziekenhuisvoorzieningen. Signaleringsrapport thermische behaaglijkheid in verpleeghuizen; 2002. Utrecht. [8] VIPA. Evaluatiecriteria ecologisch bouwen; Vlaams Infrastructuurfonds voor Persoonsgebonden Aangelegenheden; 2002. Brussel. [9] ISO/TS 14415. Ergonomics of the thermal environment-application of international standards to people with physical disabilities. Switzerland: International Organization for Standardization; 2005. [10] Verheyen J. Thermal comfort requirements for patients, staff and visitors in healthcare facilities. In: Proceedings of 8th Belgian Passive House Conference, 11e13 September 2009, Brussels; p. 320e331. [11] Fanger PO. Thermal comfort e Analysis and applications in environmental Engineering. Copenhagen: Danish Technical Press; 1970. [12] de Dear R, Brager GS. Developing an adaptive model of thermal comfort and preference. Berkeley, California: Centre for the Built Environment; 1998. [13] ISO 7730. Ergonomics of the thermal environment e Analytical determination and interpretation of thermal comfort using calculation of the PMV and PPD indices and local thermal comfort criteria. Geneva, Switzerland: International Organization for Standardization; 2006.
[14] ASHRAE Standard 55-2004. Thermal environmental conditions for human occupancy. Atlanta, GA: American Society of Heating, Refrigerating, and Airconditioning Engineers; 2004. [15] Cen EN 15251. Indoor environmental input parameters for design and assessment of energy performance of buildings addressing indoor air quality, thermal environment, lighting and acoustics. Brussels: European Committee for Standardization (CEN); 2007. [16] Hill LD, Webb LH, Parsons KC. Carers’ view of the thermal comfort requirements of people with physical disabilities. In: Proceedings of the IEA 2000/ HFES 2000 Congress, San Diego; 2000. [17] Parsons KC. The effects of gender, acclimation state, the oportunity to adjust clothing and physical disability on requirements for thermal comfort. Energy and Buildings 2002;34:593e9. [18] Webb LH, Bailey AD, Parsons KC. A software tool that provides guidance on thermal comfort conditions for people with physical disabilities. In: Proceedings of the IEA 2000/HFES 2000 congress, 2000, San Diego: p. 708e711. [19] Yoshida JA, Banhidi L, Polinszky T, et al. A study on thermal environment for physically handicapped persons e Results from Japanese-Hongarian joint experiment in 1990. Journal Thermal Biology 1993;18(5/6):363e75. [20] Webb LH, Parsons KC. Thermal comfort requirements for people with physical disabilities; Sustainable Buildings. In: Proceedings of the BEPAC and EPSRC mini conference, Oxford; 1997. [21] Khodakarami J, Knight I. Measured thermal comfort conditions in Iranian hospitals for patients and staff. In: Proceedings of clima 2007: wellbeing indoors, Helsinki; 2007. [22] Khodakarami J, Knight I. Thermal comfort requirements in Iranian hospitals. In: Proceedings of Clima 2007: wellbeing indoors, Helsinki; 2007. [23] Skoog J, Fransson N, Jagemar L. Thermal environment in Swedish hospitals summer and winter measurements. Energy and Buildings 2005;37:872e7. [24] Fransson N, Västfjäll D, Skoog J. In search of comfortable indoor environment: a comparison of the utility of objective and subjective indicators of indoor comfort. Building and Environment 2007;42:1886e90. [25] Hwang RL, Lin TP, Cheng MJ, Chien JH. Patient thermal comfort requirement for hospital environments in Taiwan. Building and Environment 2007;42:2980e7. [26] Yau YH, Chew BT. Thermal comfort study of hospital workers in Malaysia. Indoor Air 2009;19:500e10. [27] McCullough EH, Zbikowski PJ, Jones BW. Measurement and prediction of the insulation provided by bedding systems. ASHRAE Transactions 1987;93(Part 1):1055e68. [28] Voelker C, Hoffmann S, Kornadt O, Arens E, Zhang H, Huizenga C. Heat and moisture transfer through clothing. In: Proceedings of the eleventh International IBPSA Conference, Glasgow; 2009. [29] Lin Z, Deng S. A study on the thermal comfort in sleeping environments in the subtropics e Developing a thermal comfort model for sleeping environments. Building and Environment 2008;43:70e81. [30] Lin Z, Deng S. A study on the thermal comfort in sleeping environments in the subtropics e Measuring the total insulation values of the bedding systems commonly used in the subtropics. Building and Environment 2008;43:905e16. [31] ASHRAE. Handbook of fundamentals; chapter 8: thermal comfort. Atlanta, GA: American Society of Heating, Refrigerating, and Air-conditioning Engineers; 2005. [32] ISO 10551. Ergonomics of the thermal environment e Assessment of the influence of the thermal environment using subjective judgement scales. Geneva, Switzerland: International Organization for Standardization; 2001.
Contents lists available at ScienceDirect
Building and Environment journal homepage: www.elsevier.com/locate/buildenv
Thermal comfort of patients: Objective and subjective measurements in patient rooms of a Belgian healthcare facility Jan Verheyen a, *, Nele Theys a, Luk Allonsius a, Filip Descamps b,1 a b
University College of Antwerp, Department of Applied Engineering & Technology, Construction Engineering, Paardenmarkt 92, 2000 Antwerpen, Belgium Vrije Universiteit Brussel, Department of Architectural Engineering, Pleinlaan 2, 1050 Brussel, Belgium
a r t i c l e i n f o
a b s t r a c t
Article history: Received 17 September 2010 Received in revised form 27 November 2010 Accepted 11 December 2010 Available online 21 December 2010
In healthcare facilities, the prediction of mean thermal comfort perception of patients and staff is necessary to formulate requirements for the architectural and building systems design and control, and for establishing guidelines for the use of clothing and bedding systems. In this study thermal comfort of patients is evaluated by comparing objective (environmental and personal) parameters and subjective measurements (questionnaires) of thermal comfort for different groups of patients, according to the ward they are occupied in. The study involved 99 patients of maternity, oncology, neurology, gastro-enterology, abdominal surgery and thoraco-vascular surgery wards. T tests reveal no significant difference between Predicted Mean Vote (PMV) obtained from objective measurements and Actual Mean Vote (AMV) for all the different wards except for neurology. Binomial tests show that the difference between the predicted percentage of dissatisfied (PPD) obtained by application of the PPD-formula in ISO 7730 as function of Actual Mean Vote and PPD obtained from personal acceptability votes is not significant on a 5% level for all wards. This means that PMV and PPD indices may be used to adequately predict mean thermal responses for these wards except for neurology. Ó 2010 Elsevier Ltd. All rights reserved.
Keywords: Thermal comfort Patients Hospital Healthcare facility
1. Introduction Design and operation of patient rooms should primarily aim at providing a healthy and healing environment for the patients recovering from surgery, injury or disease. There has been growing scientific evidence that the physical environment has an impact on health and well-being [1]. Every physiological strain applied to the patient will induce extra stress on top of stress related to the disease or injury of the patient which is undesired unless medical treatment requires so. The thermal environment can also be an important source of undesired physiological strain on the body. It is therefore necessary to know how patients perceive thermal comfort. Thermal comfort of patients can be different from that of the healthy population because the nature of the physical disability will affect thermophysiology, thermal sensation, metabolism, blood flow, regulatory response, such as vasomotor control of body skin temperature or the ability to sweat. Depending on the disability and health status of a patient, the adaptive opportunity may be
* Corresponding author. Tel.: þ32 (0) 494 9113 52 (Mobile); fax: þ32 (0) 3 23186 70. E-mail address: [email protected] (J. Verheyen). 1 Daidalos Peutz bouwfysisch ingenieursbureau. 0360-1323/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.buildenv.2010.12.014
restricted, limiting the range of thermal environmental conditions providing comfort. Also the use of technical aids (such as wheelchairs), medical treatment and the use of drugs can affect thermal sensation and comfort perception. The aged are more largely represented in hospitals in Belgium than they have part in the overall demographic composition. It is generally considered that the ability to thermoregulate decreases with age [2] and that older people prefer slightly higher temperatures [3] although the latter can be attributed to lower activity levels and clothing differences of the elderly compared to their younger counterpart.
1.1. Guidelines and recommendations for thermal environmental conditions of patient rooms Some recommendations and guidelines for design of patient rooms for healthcare inpatient nursing are summarized in Table 1. According to Ref. [5] each room should have individual temperature control. Lower or higher temperatures than the ones prescribed in the standard shall be permitted when patients’ comfort and (or) medical treatment require those conditions [4]. Thermal comfort requirements should be considered on an individual basis for people with physical disabilities [3]. This is confirmed in the international standard for thermal comfort for
1196
J. Verheyen et al. / Building and Environment 46 (2011) 1195e1204
Table 1 Thermal comfort ranges in some standards and guidelines. RH [%]
Design temperature [ C]
Maximum air velocity [m/s]
ASHRAE; 2008 [4] ASHRAE; 2007 [5]
Max. 60 30 (W)e50 (S)
21e24 21e24
e e
ISIAQ; 2003 [6] College Bouw ziekenhuis-voorzieningen; 2002 [7] VIPA; 2002 [8]
<60%, preferable <50% e
e 22.5e25.5
e e
40e65 30e70 30e70
23.5e25.5/21e23 23e26/20e24 22e27/19e25
0.18/0.15 0.22/0.18 0.25/0.21
Inpatient nursing: patient room Hospital and outpatient facilities: nursing: patient room Healthcare: patient rooms Uses 0.5 < PMV < 0.5 for acceptable comfort range Optimal (summer/winter) Normal (summer/winter) Minimal (summer/winter)
(W, S) ¼ Winter, Summer.
people with special requirements [9], stating that consideration of individual requirements may be even more necessary for the disabled than for standard persons. Furthermore, the inclusion of personal control possibilities is recommended [10].
1.2. Previous research For evaluation of thermal comfort and prescriptions of design requirements, models based on Fanger [11] are worldwide most frequently used in current standards to assess thermal comfort, in some cases improved by an adaptive thermal comfort model for naturally ventilated buildings [12]. These models, however, are only applicable to healthy adults in moderate thermal environments, because they were based on laboratory and field studies involving mainly college-aged students in moderate environments. ISO 14415 [9] is meant for application in addition to ISO 7730 [13] for determination and interpretation of thermal comfort in the case of people with special requirements, such as people with physical disabilities, pregnancy, sickness or normal aging. This standard contains general considerations concerning thermal comfort perception as well as specification of conditions for thermal comfort. Also changes in thermophysiology and thermosensation, thermoregulatory responses or thermal comfort perception are discussed and summarized for some specific disabilities in relation to thermal environments of concern. 1.2.1. Adaptive thermal comfort and patients An adaptive thermal comfort theory was developed as a result of extensive field studies that revealed non-universality of the applicability of traditional thermal comfort theory based on laboratory experiments. Adaptation can be interpreted broadly as the gradual diminution of the organism’s response to repeated environmental stimulation. More specific, adaptation can be categorized in three main groups [12]: behavioral adjustment (personal, technological and cultural responses that change the heat and mass flows reigning the thermal balance of the human body), physiological
Table 2 Total clothing insulation IT [Clo] taken from Ref. [30]. Ac (%) Percentage of body surface area coverage: 23.3 48.0 59.1 67.0 94.1 M1 M1 M1 M1 M1 M1 M1 M1 M1
þ þ þ þ þ þ þ þ þ
Q1 þ S1 Q2 þ S1 B þ S1 Q1 þ S2 Q2 þ S2 B þ S2 Q1 Q2 B
1.57 1.57 1.57 1.38 1.38 1.38 0.98 0.98 0.98
2.15 1.84 1.82 1.65 1.53 1.43 1.16 1.14 1.07
2.72 2.24 2.08 2.15 1.93 1.76 1.43 1.42 1.24
2.88 2.41 2.18 2.62 2.20 1.80 1.90 1.69 1.45
4.56 3.73 2.56 4.34 3.55 2.40 4.03 3.03 2.11
adaptation, resulting from exposure to thermal environments in diminution of the strain induced by this exposure and psychological adaptation, including the altering of perception of sensor signals due to past experiences and future expectations that in turn alters reaction of the human body in order to maintain thermal equilibrium. The PMV/PPD is a static heat balance model that can adequately predict mean thermal sensation votes for indoor environments that occupants have no or limited control over, such as centrally controlled HVAC buildings. It can be seen as a partially adaptive thermal comfort model since it incorporates the effects of some behavioral adaptations to thermal environment like clothing and air velocity adjustments. Adaptive thermal comfort theory incorporates all effects of the opportunity to adapt. It allows a broader range for indoor acceptable environments as function of outdoor thermal environmental characteristics and is optional in ASHRAE STD 55 [14], ISO 7730 [13] and EN 15251 [15] for occupant controlled, naturally conditioned buildings with operable windows in warm periods or warm climatic zones. People with physical disabilities have restricted adaptive opportunity and special attention should be paid to this user group especially in conditions away from thermal neutrality as uncomfortable conditions affect patients both physically and mentally [16]. Patients may have restricted mobility and the ability to thermoregulate by behaving appropriately (e.g. move out of non comfortable environment or adjust clothing) may be severely restricted. Therefore the adaptive capability should be assessed when designing for comfortable indoor environment [17]. 1.2.2. Laboratory studies A reliable method to obtain ranges of acceptable thermal comfort conditions, is to conduct thermal comfort enquiries in combination with objective measurements of physical environmental and personal parameters of a group of people representative for total population of interest in a controlled climatic chamber. For patients with particular disabilities, these tests were conducted by Parsons and Webb [3,17,18] and by Yoshida et al. [19]. Of course, for reasons of medical treatment, these experiments cannot be conducted for (a representative group of) all patients. In an exploratory laboratory experiment, Webb and Parsons compared objective and subjective measurements of 16 persons with and without physical disabilities [20]. The disabilities concerned included cerebral palsy, spina bifida, stroke, Friedrich’s ataxia, blindness, paralysis, heart condition, encephalitis, Guillain-Barré syndrome, missing limbs and metal in legs. The results of laboratory experiments show no significant differences (p < 0.05) between mean responses of subjects with and without physical disabilities. Variation of responses was greater in cool to neutral conditions and smaller for slightly warm to warm conditions. The authors concluded
J. Verheyen et al. / Building and Environment 46 (2011) 1195e1204
that thermal comfort requirements for people with physical disabilities should be considered on an individual basis. Results of the main laboratory studies were integrated into a software tool that allows to determine the thermal comfort requirements of people with physical disabilities [18]. General guidance is provided in terms of a summary of findings as well as mean and variation in responses and the possibility to interrogate individual responses. The software incorporates results from people with cerebral palsy, spinal injury, accident including tetra- and quadriplegia, spinal degenerative conditions, spina bifida, hemiplegia including stroke, polio, osteo-arthritis, rheumatoid arthritis, head injury and multiple sclerosis. The laboratory results show that ISO 7730 [13] can be used without modification as a first approximation to the average responses of people with and without physical disabilities. The variation in responses is different from that of the healthy population and individual characteristics should be considered when assessing thermal environment or determining requirements for people with physical disabilities. 1.2.3. Field studies Apart from these laboratory studies, several field studies were conducted concerning thermal comfort of people with different physical disabilities, using different methods. Hill et al. [3,16] conducted field studies of people with physical disabilities and performed enquiries of carers and professionals showing that functional abilities diminish for people with physical disabilities when thermal environments are not correctly targeted at their needs, which may lead to discomfort, increased pain and poorer quality of life. In 4 Iranian hospitals, objective measurements of thermal comfort perception of patients in 14 hospital rooms are compared with requirements of thermal comfort in standards [21]. In another study, the possibilities to reconcile the different expectations towards thermal comfort of patients and staff in the same room are compared [22]. The authors suggest different zones for patients and for staff because of different thermal comfort requirements [22]. This can be achieved by differentiating radiant temperature by application of directional radiant heating via warm surfaces or ceiling for patients and cooling via the floor for staff. Attention on ventilation is needed to provide low absolute humidity levels to avoid condensation on the cooling floor. In an orthopaedic ward in a Swedish hospital the thermal environment was analyzed using objective measurements in combination with subjective measurements. Forty non-bedridden patients and 35 staff members were questioned in winter and summer season [23]. The difference between patient and staff thermal comfort perception is greater in winter than it is in summer despite similar temperatures in both seasons. For both seasons considered separately, calculated values of temperature for acceptable thermal comfort differ for patients and staff, in spite of both groups indicating to accept the thermal environment in the questionnaires. It was concluded that staff and patients cannot be treated as one coherent group of users with the same needs and preferences concerning indoor environment. The same data, but pooled for patient and staff population, were used to investigate the usability of objective and subjective indicators of indoor overall comfort by multiple regression analysis with physical and rated indicators apart and together [24]. Subjective sensory ratings were significantly better than objective indicators at predicting overall rated indoor comfort. The best prediction result of overall rated indoor comfort was obtained using a combination of subjective measures and physical variables. In Taiwan, one of the most extensive field studies of thermal comfort conducted in a hospital environment concerning patient thermal comfort perception is conducted [25]. A comparison of
1197
thermal environment of the hospital was made with ASHRAE standard 55 recommendations on one hand and a comparison with subjective measurements on the other hand. Probit regression analysis resulted in preferred effective temperature (ET*) limits for the patient population. Based on chi-square tests, the authors concluded that physical strength has a highly significant effect on thermal sensation, especially in winter, but gender, age and acclimatization have not. Patients expected a warmer indoor environment than healthy people did. Preferred and neutral temperature for patients were 23.9 and 22.9 C ET* in winter and 24.6 and 24.0 C ET* in summer. The frail wanted a neutral temperature of 0.3 C ET* higher than that of the vigorous population and there was a 1.5 and 0.8 C difference in preferred ET* for winter and summer respectively. Regression model of AMV of patients on ET* indicated that patients thermal sensitivity is blunt. A study of thermal comfort was conducted in four hospitals in tropical Malaysia involving a total of 114 occupants [26]. Subjective and objective measurements were conducted simultaneously. However, only staff was involved in answering to the questionnaires. PMV calculated from objective measurements was compared to AMV of staff. The authors concluded that the significant deviation between AMV and PMV strongly implied that PMV method might not be suitable for application in hospitals in the tropics. 1.2.4. Different hospital users Studies have been done on the thermal comfort requirements for different hospital users [7,10,22]. This is done by considering the differences in thermal insulation of clothing and activity levels of different user groups. The influence of disease, treatment or other influence specific to the hospital setting is neglected. Results show the difficulty to reconcile the needs concerning the thermal environment of the different hospital users. Most field studies do not differentiate between patients as function of the type of disability or treatment taken, but consider patients and staff together [24], patients and staff apart [23], staff only [26] or patients only [25]. Other studies focus on the thermal comfort requirements for people with specific disabilities. For a review of the latter, the reader is referred to Parson’s book on human thermal environments [3], in which a complete chapter is dedicated to thermal comfort for special populations including people with physical disabilities. However no study is known by the authors that focuses on thermal comfort of different groups of hospital patient population according to the ward or department this population is occupied in.
1.3. Objective This study concentrates on different groups of patients, sorted per department or ward they are occupied in. Reason for this is of a practical nature; designing for comfortable indoor environment could be specifically diversified per department or ward. The nature of the disability will determine requirements, as the physical disability may influence thermal sensation and perception. To address the specific thermal comfort needs of patients, the design of the bed, the room and the control measures for the thermal environment per department or ward can be diversified. This approach is considered useful because patients are grouped in specific wards depending on the disease or disability they are receiving treatment for. It is intended to provide a first indication of the usability of the Fanger PMV (Predicted Mean vote) and PPD (Predicted Percentage of Dissatisfied) indices for groups of patient population per ward (according to the surgical recovery or medical treatment they are obtaining).
1198
J. Verheyen et al. / Building and Environment 46 (2011) 1195e1204
Fig. 1. Ac: percentage of body surface area coverage taken from [30].
2. Research method 2.1. Thermal comfort and bedding For the determination of the clothing insulation of patients, only few studies take bedding insulation into account [7,10,22]. In the former study [7] it is not indicated how total clothing insulation values for bedding systems (taken from measurements performed by McCullough et al. [27]) were transformed for input as clothing insulation values in PMV calculation according to ISO 7730 [13]. Estimated clothing insulation values have large influence on the results of the calculation of acceptable thermal comfort [10]. A comprehensive description of a model to account for heat and moisture transfer through clothing is described by Voelker et al. [28]. Global thermal comfort requirements for different hospital users in different situations (defined by clothing insulation ranges for summer and winter conditions and activity level ranges) including patients in bed are described in earlier research [10]. A method to calculate PMV and PPD for sleeping environments [29] is based on the thermal comfort index of Fanger where PMV is related to the imbalance between the actual heat flow from the human body and the heat flow required to obtain thermal comfort in the same environmental conditions. This imbalance is referred to as the thermal load and results in the PMV value by multiplying it with a sensitivity coefficient. This sensitivity is assumed to be valid also for sleeping conditions. The relation between PMV and PPD was also assumed by Lin and Deng to remain valid for sleeping conditions. The application of the method described in ISO 7730 is recommended to be limited to metabolic rate levels between 46 W/m2 and 232 W/m2 (0.8 and 4.0 Met). Application by Lin and Deng assumes that the same model for thermal comfort prediction can be used for activity levels of 40 W/m2 (0.7 Met), resembling metabolic heat production during sleeping activity. Lin and Deng formulate the heat balance equation at the skin surface in terms of operative temperature instead of mean radiant temperature and air temperature separately. The energy balance equation of Fanger’s model is adapted for heat and moisture transfer from the skin to the environment by taking the total clothing insulation resistance into account. The total insulation resistance of a bedding ensemble is measured for different bedding systems in combination with sheets and clothing worn [30] and includes the resistance of the air layer at the surface of the bedding
system and the human body at locations not covered. The air layer resistance is not included in the clothing insulation value used in the ISO 7730 method. The total insulation resistance represents a resistance to heat (or moisture) transfer uniformly spread over the entire body surface resulting in the same total heat (or moisture) transfer as the non-uniformly spread ensemble of bedding system and clothing worn. For moisture transfer, one value for the total vapor permeation efficiency (im ¼ 0.38) was considered to be reasonably accurate, also for bedding systems. Together with the Lewis ratio (LR, being he/hc; division between the evaporative (he) and convective (hc) surface transfer coefficients) and the total insulation resistance to heat transfer (Rt), this provides an approximation of the total resistance to moisture transfer (Re,t) through the following equation: im $ LR ¼ Rt/Re,t [31]. Skin wettedness w (being the evaporative heat loss from the skin (Esk) divided by the maximum possible evaporative heat loss (Emax) [31]), was chosen to be 0.06, which corresponds to no regulatory sweating for normal conditions. Evaporative heat transfer occurs only due to moisture diffusion through the skin, which holds true for M W (metabolic heat production work delivered) smaller or equal to 1 Met or 58.2 W/m2. Comfort equation proposed by Lin and Deng for sleeping environments [29]:
tsk;req t0 0:06im LR psk;s pa MW ¼ þ þ 0:014Mð34 ta Þ Rt Rt þ 0:0173Mð5:87 pa Þ
where M ¼ metabolism [W/m2] W ¼ work delivered [W/m2] tsk,req ¼ mean skin temperature, required for thermal comfort [ C] ¼ 35.7e0.028(M W) to ¼ operative temperature [ C] Rt ¼ total resistance to heat transfer of a bedding system including air layer around (covered) body im ¼ total vapor permeation efficiency [ND] ¼ 0.38 LR ¼ Lewis ratio [K/kPa] ¼ 16.5 K/kPa psk,s ¼ saturated water vapor pressure at tsk [kPa] pa ¼ ambient water vapor pressure [kPa] ta ¼ ambient air temperature [ C] PMV can be derived from the comfort equation using the following relationship [29]:
J. Verheyen et al. / Building and Environment 46 (2011) 1195e1204
1199
Table 3 Bedding coverage in current study, related to Table 2. Bedding garment [current study]
Sheet
Comforter single
Sheet þ blanket/Melton
Comforter double
Related to Table 2
B
Q1
Q2
Q3
PMW ¼ ½0:303 expð0:036MÞ þ 0:028L ¼ aL where L ¼ thermal load on the body ¼ difference between left hand side and right hand side of comfort equation a ¼ sensitivity coefficient Laboratory measurements [27,30] provide values for the total insulation resistance of bedding systems including the air layer around a covered bed for different bedding systems occupied by a thermal manikin in supine position on a single bed. The bedding systems measured in the studies consist of different combinations of quilt or blanket, bed, mattress and sleepwear, in different configurations of body surface area coverage. Measurements by Lin and Deng were conducted maintaining a fixed manikin surface temperature in an environmental chamber with air temperature and mean radiant temperature of 22 C, relative humidity of 50% and air velocity 0.15 m/s. Measurements by McCullough were conducted maintaining fixed manikin surface temperatures and constant heating power of the manikin (by altering environmental chamber temperature.) Air velocity was kept below 0.1 m/s. The result of the nude manikin placed horizontally in the chamber gave the same result for both studies; 0.73 Clo (¼0.113 m2 K/W) (Fig. 1). In the current study total insulation values of bedding systems were estimated by visual inspection relating to the detailed descriptions and photos of mattress (only one type was included and is referred to as M1), sheets and blankets as well as sleepwear and body surface area coverage in the original publication of Lin and Deng [30]. In Table 4, S1 refers to full-slip sleepwear and is described as long-sleeved, 100% cotton with thickness 0.46 mm, whereas S2 refers to half-slip sleepwear specified as short-sleeved, 100% cotton with thickness 0.83 mm in the original work by Lin and Deng [30]. 2.2. Data collection Objective (physical) and subjective measurements of thermal comfort were gathered for different groups of hospital patient population in a clinical healthcare facility in Belgium (climate: temperate; mild winters, cool summers; rainy, humid, cloudy) in March 2009, in-between winter and springtime. Only patients in single and double rooms were involved. The study includes 99 patients of different wards (see Table 5) for which local ethical committee approval was obtained.
Table 4 Clothing insulation characteristics of patients [30,31]. Description of clothing used for visual observation
[Clo]
Related to Table 2 in case of patient in bed
Sleeveless short gown (thin) Sleeveless long gown (thin) Short-sleeved hospital gown Short-sleeved short robe (thin) Short-sleeved pajamas (thin) Long-sleeved long gown (thick) Long-sleeved, short wrap robe (thick) Long-sleeved pajamas (thick) Long-sleeved, long wrap robe (thick)
0.15 0.20 0.31 0.34 0.42 0.46 0.48 0.57 0.69
S2 S2 S2 S2 S2 S1 S1 S1 S1
Patients were all individually interviewed in their room based on a questionnaire according to ISO 10551 [32] extended with other questions, considered relevant for the research project. Supplementary to the basic questions suggested in ISO 10551, questions about the satisfaction concerning indoor air quality, acoustic quality, light and visual comfort, the stability of the thermal environment, the possibility to adapt and health status were added. Starting prior to the interview, air temperature, operative temperature, relative humidity and omnidirectional air velocity were measured at a height of 1.10 m in a zone of 1 m around the patients bed with a Brüel & Kjær indoor climate analyzer model number 1213. Personal factors; activity level and clothing insulation level were estimated based on answers to the questionnaire and on observation. Clothing and bedding thermal insulation values were derived from ASHRAE standard 55 [14] and measurements by Lin and Deng [30]; see Tables 2e4. PMV and PPD were calculated according to ISO 7730 [13] and to the method developed by Lin and Deng [29] (the latter in the case of patient lying in a bed). These measurements are referred to as objective measurements. The results of the objective measurements are then used as a reference indicating the predicted mean thermal sensation responses of healthy adults. 2.3. Data analysis Two-sample t tests were conducted for thermal sensation to determine whether the difference between the mean of PMV obtained from objective and from subjective measurements is significant on a 5% level. Binomial tests were used to reveal whether the difference of the means of predicted percentage of dissatisfied obtained from objective and subjective measurements was significant on a 5% level. The objective PPD is obtained from the mean of subjective thermal sensation votes by calculation of PPD according to ISO 7730 [13]. The subjective PPD is calculated from personal acceptability votes. These data analysis methods are in agreement with the methods described in ISO 10551 [32]. 3. Results and discussion 3.1. Personal parameters Observations of total thermal insulation IT and clothing thermal insulation Icl and activity level of patients are represented by boxplots in Fig. 2. Almost half of all patients (46/99) had been either reclining or sleeping in bed during the hour prior to enquiry. Almost one-third of all patients (31/99) was just sitting or reading or writing in Table 5 Departments, wards and number of patients involved. Department Maternity Medicine
Surgery Total
Ward
Oncology Neurology Gastro-enterology Abdominal surgery Thoraco-vascular surgery
Number of patients involved 17 17 17 14 17 17 99
1200
J. Verheyen et al. / Building and Environment 46 (2011) 1195e1204
Fig. 2. Activity level and bedding and/or clothing insulation of patients.
100% 90%
maternity (mean 0.06)
80%
medicine oncology (mean -0.24) medicine neurology (mean -0.24)
70%
medicine gastro-enterology (mean -0.14)
60%
abdominal surgery (mean -0.65)
number 50% of votes
thoracovascular surgery (mean -0.35) total (mean -0.26)
40% 30% 20% 10% 0% very weak
weak
neutral
fit
very fit
Self-described health status Fig. 3. Self-described health status.
a chair, the rest was walking about and one patient had been mostly standing, relaxed during the last hour. This distribution in activity level was similar for all wards, but slightly different for oncology and abdominal surgery wards; only one of all oncology patients had been walking about and only 5 out of 17 abdominal surgery patients had been either reclining or sleeping in bed . The insulation provided by clothing and bedding system is represented by the boxplot in the middle of Fig. 2. The insulation provided by clothing of the patients that are standing or sitting in a chair is represented by the right hand side part of Fig. 2. The difference between parameters total insulation IT and clothing insulation Icl is explained in Section 2.1 of this document. In case of a patient sitting in a chair, the insulation characteristics of the chair were neglected. A sitting posture results in a decreased insulation of clothing due to compression of air layers in the clothing [14]. This
decrease may be offset by the insulation provided by the chair. For many chairs the net effect of sitting is a minimal change in clothing insulation. For this reason, ASHRAE standard 55 recommends that no adjustment to clothing insulation is to be made if there is uncertainty as to the type of chair [14]. Since both PMV/PPD methods are valid only for steady state conditions, a supplementary question was added in the questionnaire, enabling exclusion from the results when substantial altering of thermal conditions of the environment in the previous hour was indicated by the patients’ answer. No altering of thermal environment was reported in the questionnaires. Physical strength (answers to questions: do you feel very, moderate, slightly frail or vigorous) has a highly significant effect on a 5% level (p ¼ 0.000) on thermal sensation [25]. Therefore, the authors found it useful to conclude a question about the health
Table 6 Results and environmental parameters for patients grouped per category of self-described health status. Self-described health status
Number of observations
Mean PMVobj
Mean PMVsubj
STD PMVobj
STD PMVsubj
Mean operative temperature [ C]
STD operative temperature [ C]
Mean Relative Humidity [%]
STD Relative Humidity [%]
Very weak (2) Weak (1) Neutral (0) Fit (þ1) Very fit (þ2)
7 33 40 17 2
0.32 0.76 0.40 0.58 0.49
0.57 0.00 0.15 0.53 2.00
1.27 0.99 1.13 0.95 0.68
1.72 1.22 0.83 1.07 0.00
24.0 23.4 23.7 23.3 24.1
0.9 0.9 1.0 0.5 2.3
37.0 36.4 35.7 36.1 33.5
3.8 4.4 4.5 3.1 10.6
STD ¼ standard deviation.
J. Verheyen et al. / Building and Environment 46 (2011) 1195e1204
1201
100% maternity (mean 1.12)
90%
medicine oncology (mean 0.71)
80%
medicine neurology (mean 0.94)
70%
medicine gastro-enterology (mean 0.71) 60% abdominal surgery (mean 0.29) number 50% of votes 40%
thoracovascular surgery (mean 0.88) total (mean 0.78)
30% 20% 10% 0% very unsatisfied unsatisfied
neutral
satisfied
very satisfied
possibilities to adapt Fig. 4. Satisfaction with possibilities to adapt to thermal environment.
status of the patients in the enquiry. Results of self-described health status are represented in Fig. 3. Fig. 3, depicting percentages of answers in the categories of selfdescribed health status shows a clear deviation to the negative side (weak and very weak), which is to be expected in a healthcare facility. Table 6 contains mean and standard deviation of objective PMV and subjective PMV together with environmental data of operative temperature and relative humidity of the rooms of all the patients grouped per self-described health status category. Results of linear
regression of subjective PMV on self-described health status show the correlation coefficient squared to be 0.066, indicating that there is indeed an effect of influence that cannot be ignored. However, Table 6 shows that patients of one group voting in the same category of health status did not experience the same physical conditions as another group. Mean objective PMV for different groups is between 0.32 and 0.76. Nevertheless, if one would wish to develop a method for determining the PMV for overall patient population and is looking for parameters of influence to predict the mean thermal sensation vote, one could consider incorporating the
Fig. 5. Operative temperature (To), relative humidity (RH) and air velocity in hospital patient rooms.
Fig. 6. Acceptable thermal environment for all measurements.
1202
J. Verheyen et al. / Building and Environment 46 (2011) 1195e1204
Table 7 Results of objective and subjective measurements, two-sample t tests and binomial tests. 5% level of significance
Number of Mean Mean STD STD Difference between PMVobj Mean Mean Difference between observations PMVobj PMVsubj PMVobj PMVsubj and PMVsubj PPDobj PPDsubj PPDobj and PPDsubj* objective or (binomial test) subjective
Maternity Maternity Medicine Oncology Neurology Gastro-enterology Surgery Abdominal surgery Thorovascular surgery
17 17 17 14 17 17
0.30 0.58 0.87 0.58 0.45 0.52
0.29 0.06 0.06 0.21 0.12 0.41
0.93 1.00 1.09 1.01 0.94 1.35
0.99 1.20 1.09 1.63 0.93 1.00
Not significant (p ¼ 0.082) Not significant (p ¼ 0.101) Significant (p ¼ 0.038) Not significant (p ¼ 0.135) Not significant (p ¼ 0.085) Not significant (p ¼ 0.787)
6.8 5.1 5.1 6.0 5.3 8.5
5.9 12.0 0.0 7.1 0.0 5.9
Not Not Not Not Not Not
significant significant significant significant significant significant
(p (p (p (p (p (p
¼ ¼ ¼ ¼ ¼ ¼
0.680) 0.212) 0.413) 0.576) 0.396) 0.568)
*PPDsubj calculated from personal acceptability votes. STD ¼ standard deviation.
effect of health status in a model for PMV. If very weak, weak, neutral, fit and very fit would respectively correspond to 2, 1, 0, 1 and 2, the mean self-described health status has a value of 0.26; in between weak and neutral. There is however no indication that the categories used are psychologically interpreted to have the same distance from each other. There is one population group that reported a positive mean self-described health status; namely maternity. Abdominal surgery shows most negative of mean values. The mean for each group is indicated in the legend of Fig. 3. Fig. 4, shows the satisfaction (on a five-point scale) of the patients concerning possibilities to adapt to the thermal environment when they were having too cold or too hot. As examples, opening windows, changing temperature, altering bedding or clothing, sunblinds, ventilation,., were included in the questionnaire. Overall patients were satisfied with their possibilities to adapt to the thermal environment. If 2 is very unsatisfied, 1 is unsatisfied, 0 is neutral, þ1 is satisfied and þ2 is very satisfied, overall patients mean is 0.78; in between neutral and satisfied. The mean for each group is indicated in the legend of Fig. 4. There is however no indication that the categories used are psychologically interpreted to have the same distance from each other. Apart from their satisfaction, 68% of patients indicated they were able to adjust to the thermal environment by altering clothing or bedding by themselves and needed no help from nursing staff. This means 32% needed help to do so. 54% indicated being able to adjust to thermal environment by other measures (like opening windows, heating, cooling,.) without help from nursing staff. Most self-control in possibility to adapt by altering clothing or bedding as well as by other measures was found in maternity patients and least in abdominal surgery patients. Not a single patient indicated not being able to adjust. However, the need for help means an extra obstacle in achieving the adjustment. The patient has to rely upon judgement and assistance of a third person. First of all the posing of
the question implies admitting insufficient self-control and independence. Secondly, the nursing staff at service should be convinced of the validity of the complaint before taking action to adjust. Very likely the perception of thermal comfort of staff is different from that of the patient [10]. In the case of thermal discomfort these elements can lead to a delay or even cancellation of the adjustment and risks the patient of drifting further away from comfortable conditions. Nevertheless, patients are rather satisfied with their possibilities to adapt, with a deviation towards neutral for abdominal surgery patients. Maternity patients show to be most satisfied with their possibilities to adapt (Fig. 5). 3.2. Indoor environmental parameters The measurements of indoor parameters operative temperature and relative humidity in patient rooms are compared with design parameters recommended by ASHRAE and VIPA (Table 1). Temperature in the rooms is high, certainly for winter period, and some rooms fail to meet the lower limit for relative humidity level compared to VIPA and ASHRAE recommendations (Table 1). The rooms with lowest relative humidity and highest operative temperature are mainly oncology rooms. Lowest relative humidity levels indoors occur in wintertime when no humidification is provided. Air velocity for optimal design level is limited to below 0.15 m/s in wintertime according to VIPA [8]. Only two measurements (of a total of 99) trespass this limit and only one measurement (0.5 m/s) is higher than the maximum value for minimal comfort level of 0.21 m/ s set by [8]. Augmenting air velocity from 0.1 m/s to 0.2 m/s approximately corresponds to a decrease in operative temperature of 0.9 C or a decrease in PMV value of 0.33 (for sedentary activity (1 Met) in summer clothing (0.5 Clo) at 50% RH). The high air velocities measured do not occur in rooms with high room temperature.
Fig. 7. Boxplots of PMV obtained from objective and subjective measurements of different wards’ population.
J. Verheyen et al. / Building and Environment 46 (2011) 1195e1204
Fig. 6 shows the percentage of total measurements of thermal environments that are acceptable. The left hand side of Fig. 6 shows percentage of acceptable thermal environments based on 3 different determination methods; percentage of votes in central three categories of thermal sensation, percentage of votes of personal acceptability in category ‘acceptable’ and percentage of votes of thermal preference in category ‘want no change’. On the right hand side of Fig. 6 the percentage of environments that meet the criteria of acceptable thermal comfort as indicated in Table 1 is represented. In all cases operative temperature, relative humidity and air velocity measurement results (the latter only for VIPA recommendations) are checked with the ranges for acceptable thermal comfort. It can be seen that although 95% accepted the thermal environment they were in on a direct assessment, only 74% wanted no change. Overall, patients in this healthcare facility found the environment acceptable rather than unacceptable. The unacceptability index (percentage of votes in category ‘unacceptable’) was similar for all wards with an overall value of 5%. Only for oncology patients it was higher; 12% and for neurology and abdominal surgery wards there were no patients that judged the environment to be unacceptable. 78% voted in central three categories, which is almost the same as the number wanting no change. From objective measurements of operative temperature, relative humidity and air velocity, it can be concluded that 71% of environments comply with ASHRAE and 70% with VIPA for normal quality thermal environment in winter. Furthermore 9% of environments did not comply with VIPA minimal level of thermal environment. However, of all patients included in this study that were occupying these environments that do not meet the VIPA minimum criteria, only 22% voted outside the central 3 categories of thermal preference, only 33% indicated to want change and they all found it to be perfectly bearable. The central tendency (median) for answers to the tolerance assessment question (“Is it perfectly bearable/./unbearable?”) is for overall patient population and for all wards separately the category ‘perfectly bearable’. 11% of overall patient population judged it to be more or less difficult to tolerate with highest painfulness indices (percentage of judgements expressing ‘difficulty to bear’) in wards oncology (18%) and maternity (18%), gastro-enterology (14%) and lowest in neurology (6%), abdominal surgery (6%) and thoraco-vascular surgery (6%). Only one patient judged it to be unbearable. This patient was occupied in the maternity ward. 3.3. PMV/PPD indices Values in Table 7 for mean PMVobj, STD PMVobj and PPDobj and boxplots of all objective measurements in Fig. 7 are based on data of individual PMV value calculation by the method according to ISO 7730 [13] in case of patient not in bed and by the method developed by Lin and Deng [29] in case of patient in bed. Fig. 7 and Table 7 show that for all groups patients’ rated thermal sensation from subjective measurement is higher than the PMV obtained from objective measurements. This indicates that patients expect a lower thermo-neutral temperature in comparison to healthy adults. This is in contradiction with results obtained in a previous field study [25] conducted in a hospital be it in a very different climatic zone. Apart from outdoor climate, another reason for this difference could be found in the fact that patients voted very often in the neutral category and showed a high acceptability rate, indicating a bluntness to thermal sensation in comparison to healthy population. This was also observed and reported in previous research [25]. Another reason for the difference between the objective and subjective PMV, can be a systematic underestimation of thermal insulation of clothing and bedding system and/or of metabolic heat production, resulting in a lower objective PMV. Variance in PMV
1203
obtained from subjective measurements is slightly bigger than variance in PMV from objective measurements for all wards except for thoraco-vascular surgery. The largest difference is found for the gastro-enterology ward. Both laboratory and field studies show a wide variation in thermal responses of the disabled [9]. A bigger variance in PMV obtained from subjective measurements in comparison to those obtained from objective measurements is therefore not surprising. Two-sample t tests show that the difference between PMV from subjective measurements (AMV) and PMV from objective measurements is not significant on a 5% level for all individual ward population groups except for neurology. A second analysis was performed using the same techniques but after rejection of outliers determined with Grubbs’ test with a significance level of 5%. The conclusions remain the same. Finally, a third analysis is conducted using Welch’s t-test, assuming unequal variances for results of objective and subjective measurements. This leads to smaller number of degrees of freedom and slightly smaller t values. The results and conclusions remain the same, only with slightly different p-values. The p-values given in Table 7 correspond to Welch’s t-test results. Binomial tests reveal no significant difference between PPD obtained from subjective measurements and PPD calculated from subjective thermal sensation votes for all wards individually. This means that for wards’ populations included in this study, calculation method for PPD according to ISO 7730 [13] as function of PMV can be used to calculate the predicted percentage of dissatisfied for patient population of all wards concluded in this study. 4. Conclusions Thermal environment in this Belgian healthcare facility is acceptable for 95% of the patients (direct assessment), despite of 29% of thermal environments not complying with ASHRAE design ranges of temperature and relative humidity. This could indicate that environmental conditions for hospital patient rooms recommended by ASHRAE are too stringent for the patients included in this study or that adaptation possibly allows broader regions of recommended environmental parameters. PMV of patients can be calculated using ISO 7730 [13] and using a calculation method described by Lin and Deng [29] (the latter in case of the patient lying in bed) for all wards included in this study except for neurology ward. PMV may adequately predict mean thermal sensation for the majority of patient population for the wards considered except for neurology ward. The mean PMV of patients occupied in neurology ward is significantly different from the mean PMV of healthy population on a 5% level. Neurology ward occupies patients that need medical treatment for, or that are recovering from injury or surgery of the brain, bone marrow or nervous system. These parts are of vital importance to human thermophysiology and eregulation and it is therefore not surprising that thermal comfort perception of these patients differs significantly from that of the healthy population. Patients rated their thermal sensation systematically higher than predicted by objective measurements, indicating that thermoneutral temperature for patients is lower than for healthy, young population. This is not in agreement with results of a previous field study [25]. However only 5% of patients judged the thermal ambiance as unacceptable on a personal level. The higher mean of sensation votes could be the result of thermal sensation bluntness of patient population or of systematically underestimated clothing insulation and/or metabolic heat production resulting in an underestimated objective PMV. The difference between the PPD calculated by use of ISO 7730 [13] from PMV obtained from the questionnaires and PPD directly obtained from personal acceptability answers in the questionnaires
1204
J. Verheyen et al. / Building and Environment 46 (2011) 1195e1204
is not significant on a 5% level for all wards’ populations. PPDformula taken from ISO 7730 [13] as function of PMV can be used for patient population considered in this study. Acknowledgements This research project was conducted in collaboration with Ziekenhuis Oost-Limburg (ZOL). The authors wish to thank ZOL (especially Tilly Postelmans and employees of the quality control division and Philip Verheye of facility management division) for enabling the execution of the project and Jacques Geers of University College of Antwerp for assistance in the statistical analysis. References [1] Van den Berg AE. Health impacts of healing environments e A review of evidence for benefits of nature, daylight, fresh air, and quiet in healthcare settings. 2005. ISBN: 90 327 03447. [2] Havenith G. Temperature regulation and technology. Gerontechnology 2001;1 (1):41e9. [3] Parsons KC. Human thermal environments: the effects of hot, moderate and cold environments on human health, comfort and performance. 2nd ed. London: Taylor and Francis; 2003. [4] ASHRAE/ASHE STD 170-2008. Ventilation of health care facilities. Atlanta, GA: American Society of Heating, Refrigerating, and Air-conditioning Engineers; 2008. [5] ASHRAE. Handbook HVAC applications: chapter 7; healthcare facilities. Atlanta, GA: American Society of Heating, Refrigerating, and Air-conditioning Engineers; 2007. [6] ISIAQ. Indoor air quality for healthcare facilities. International Society of Indoor Air Quality and Climate; 2003. [7] College Bouw Ziekenhuisvoorzieningen. Signaleringsrapport thermische behaaglijkheid in verpleeghuizen; 2002. Utrecht. [8] VIPA. Evaluatiecriteria ecologisch bouwen; Vlaams Infrastructuurfonds voor Persoonsgebonden Aangelegenheden; 2002. Brussel. [9] ISO/TS 14415. Ergonomics of the thermal environment-application of international standards to people with physical disabilities. Switzerland: International Organization for Standardization; 2005. [10] Verheyen J. Thermal comfort requirements for patients, staff and visitors in healthcare facilities. In: Proceedings of 8th Belgian Passive House Conference, 11e13 September 2009, Brussels; p. 320e331. [11] Fanger PO. Thermal comfort e Analysis and applications in environmental Engineering. Copenhagen: Danish Technical Press; 1970. [12] de Dear R, Brager GS. Developing an adaptive model of thermal comfort and preference. Berkeley, California: Centre for the Built Environment; 1998. [13] ISO 7730. Ergonomics of the thermal environment e Analytical determination and interpretation of thermal comfort using calculation of the PMV and PPD indices and local thermal comfort criteria. Geneva, Switzerland: International Organization for Standardization; 2006.
[14] ASHRAE Standard 55-2004. Thermal environmental conditions for human occupancy. Atlanta, GA: American Society of Heating, Refrigerating, and Airconditioning Engineers; 2004. [15] Cen EN 15251. Indoor environmental input parameters for design and assessment of energy performance of buildings addressing indoor air quality, thermal environment, lighting and acoustics. Brussels: European Committee for Standardization (CEN); 2007. [16] Hill LD, Webb LH, Parsons KC. Carers’ view of the thermal comfort requirements of people with physical disabilities. In: Proceedings of the IEA 2000/ HFES 2000 Congress, San Diego; 2000. [17] Parsons KC. The effects of gender, acclimation state, the oportunity to adjust clothing and physical disability on requirements for thermal comfort. Energy and Buildings 2002;34:593e9. [18] Webb LH, Bailey AD, Parsons KC. A software tool that provides guidance on thermal comfort conditions for people with physical disabilities. In: Proceedings of the IEA 2000/HFES 2000 congress, 2000, San Diego: p. 708e711. [19] Yoshida JA, Banhidi L, Polinszky T, et al. A study on thermal environment for physically handicapped persons e Results from Japanese-Hongarian joint experiment in 1990. Journal Thermal Biology 1993;18(5/6):363e75. [20] Webb LH, Parsons KC. Thermal comfort requirements for people with physical disabilities; Sustainable Buildings. In: Proceedings of the BEPAC and EPSRC mini conference, Oxford; 1997. [21] Khodakarami J, Knight I. Measured thermal comfort conditions in Iranian hospitals for patients and staff. In: Proceedings of clima 2007: wellbeing indoors, Helsinki; 2007. [22] Khodakarami J, Knight I. Thermal comfort requirements in Iranian hospitals. In: Proceedings of Clima 2007: wellbeing indoors, Helsinki; 2007. [23] Skoog J, Fransson N, Jagemar L. Thermal environment in Swedish hospitals summer and winter measurements. Energy and Buildings 2005;37:872e7. [24] Fransson N, Västfjäll D, Skoog J. In search of comfortable indoor environment: a comparison of the utility of objective and subjective indicators of indoor comfort. Building and Environment 2007;42:1886e90. [25] Hwang RL, Lin TP, Cheng MJ, Chien JH. Patient thermal comfort requirement for hospital environments in Taiwan. Building and Environment 2007;42:2980e7. [26] Yau YH, Chew BT. Thermal comfort study of hospital workers in Malaysia. Indoor Air 2009;19:500e10. [27] McCullough EH, Zbikowski PJ, Jones BW. Measurement and prediction of the insulation provided by bedding systems. ASHRAE Transactions 1987;93(Part 1):1055e68. [28] Voelker C, Hoffmann S, Kornadt O, Arens E, Zhang H, Huizenga C. Heat and moisture transfer through clothing. In: Proceedings of the eleventh International IBPSA Conference, Glasgow; 2009. [29] Lin Z, Deng S. A study on the thermal comfort in sleeping environments in the subtropics e Developing a thermal comfort model for sleeping environments. Building and Environment 2008;43:70e81. [30] Lin Z, Deng S. A study on the thermal comfort in sleeping environments in the subtropics e Measuring the total insulation values of the bedding systems commonly used in the subtropics. Building and Environment 2008;43:905e16. [31] ASHRAE. Handbook of fundamentals; chapter 8: thermal comfort. Atlanta, GA: American Society of Heating, Refrigerating, and Air-conditioning Engineers; 2005. [32] ISO 10551. Ergonomics of the thermal environment e Assessment of the influence of the thermal environment using subjective judgement scales. Geneva, Switzerland: International Organization for Standardization; 2001.